I am currently developing image processing and visualisation tools to improve cardiac MR including: cardiac diffusion tensor imaging, cardiomyocyte tractography, deep learning and multi-modality image registration.
Contact me
I've developed a MATLAB image processing and analysis software tool for cardiac DTI. It outputs multiple DTI maps including: fractional anisotropy, mean diffusivity, tensor mode, helix angle and secondary eigenvector orientation in relation to the local cardiac coordinates. It includes an automatic left ventricle segmentation aided by a convolutional neural network. It is used routinely by clinicians in our department.
Automatic image classification and left ventricular myocardial segmentation with convolutional neural networks. The fully automated cardiac DTI analysis with deep learning performed effectively with high levels of accuracy when compared to an experienced user. Of note, the myocardial segmentation learned to correctly exclude papillary and right ventricular trabeculation from the LV myocardium, even in patient data with a more varied morphology.
Left ventricular tractograms of healthy human heart. The tracts are colour coded according to the local helix angle and shown for different depths of the myocardial wall. This data was acquired in vivo.
Visualisation of the entire diffusion tensor with superquadric glyphs. This particular set of images show an horizontal long-axis plane of the heart, viewed from different angles. This data was acquired in vivo.
Developed MATLAB and Python tools to automatically find and measure orientation of specific features in microscopy histology images of the heart muscle.
Thank you! Your submission has been received!
Oops! Something went wrong while submitting the form :(
